GITTA-Logo
PDF Version of this document Search Help Glossary

Lesson Navigation IconSuitability analyis

Unit Navigation IconDecision support with GIS

LO Navigation IconSuitability analysis with the help of suitability maps

LO Navigation IconSpatial Decision Support Systems (SDSS)

LO Navigation IconDecision Support with multiple criteria

LO Navigation IconDecision support for multiple objectives

LO Navigation IconDerivation of the criteria for suitability analysis

Unit Navigation IconBoolean Overlay

Unit Navigation IconWeighted overlay

Unit Navigation IconDetermining weights

Unit Navigation IconSummary

Unit Navigation IconRecommended Reading

Unit Navigation IconGlossary

Unit Navigation IconBibliography

Unit Navigation IconMetadata


GITTA/CartouCHe news:


Go to previous page Go to next page

Decision Support with multiple criteria

In the simplest case of decision support with GIS, locations or areas can be found that meet or optimize multiple criteria for one objective. In the case study of St. Gittal, the objective in a first approach is to identify the best habitat for the wolf in the municipality. Let's assume the wolf prefers areas far away from residential areas and areas covered with forest. From that, two search criteria can be formulated. If there are more than one criterion but only one objective, as in this case, it is called a multi-criteria evaluation (MCE).

Standard procedure for MCE:

  1. Define the problem: the first step of an MCE is the definition of the problem. In the case study of St. Gittal, the problem definition could be as follows: "Which parts of the municipality is suitable wolf habitat?"
  2. Select the criteria: the next step is to select the criteria. The chosen criteria should reflect the characteristics of the desired location or area as closely as possible. Criteria can be both spatial (geometry, topology) as well as factual (attributes). For example, a spatial criterion could be the distance between a potential habitat of the wolf to the nearest settlement. The restriction of the land use category "forest" on the other hand is a factual criterion. Furthermore, there are hard, "must-have" criteria and soft, "nice-to-have" criteria.
  3. Operationalization of the criteria: when the criteria are determined, they must be translated to precise, measurable indicators. This process is called operationalization. The criterion "not too close to the settlement" could be translated to the specification of a minimum distance in meters of the residential area. It is said that the criteria are to be operationalized to computable indicators. In most cases, the individual criteria correspond to one data layer in a GIS (forest layer; layer with the distance to the residential area).
  4. Creation of a common reference – data integration: Data integration creates comparability through a common measurement scale, the same data type (raster / vector), and the same resolution and reference system.
  5. Intersection: identification of the most suitable areas. Now the different criteria are allocated to find the desired location. There are several possible approaches:
    • Logical (Boolean) overlay: in each layer there is only binary true/false information (forest / non forest). By logically intersecting this information, the desired locations and areas wanted are determined. This topic is explained in the unit "Boolean overlay".
    • Weighted overlay: rarely is the simple distinction between true and false possible in the complexity of reality. A significant improvement of the results can be achieved by providing the individual data layers with weights. For example, in the case study, the distance to the settlement could be much less important than the forest as a sanctuary. A weighting factor of five could be assigned to the layer with the distance to the village. These topics are discussed separately (units "Weighted overlay" and "Determining the weights").
    • Fuzzy overlay: erroneous input data and the wrong choice of criteria can lead to evaluation errors. In a multi-criteria evaluation, appropriate areas could be ignored and unsuitable areas could be falsely classified as suitable. A solution to this problem is the cancelation of sharp boundaries. For the spatial data, this means that boundaries are not represented as sharp lines but as transition zones. With attributes, fuzzy ranges of values replace sharp class limits. This concept is based on the idea of the fuzzy set theory (fuzzy in the sense of blurred). These approaches are discussed further only in the intermediate level.
  6. Verification / evaluation: the final step involves comparing the results with a reference. This is possible if reference data, collected in the field, are available ("ground truth"). This last step is often neglected. But be aware: a suitability map without an assessment of its quality and reliability is often not worth the paper on which it was printed!

MCE Example:

Determine the evaluation criteria for the multi criteria evaluation for the TWW Project. (Click here for more information)

Top Go to previous page Go to next page